I've studied climate and its effects on life—all kinds of life ---for more than four decades, starting in 1968. Along the way, among other things, I developed a computer model of forests that in the 1990s we used to forecast the effects of climate on jack pine forests in Michigan that were the only habitat of the endangered Kirtland’s warbler. A lot of effort was going into saving the bird’s habitat, and I wondered if, with global warming, it might all be in vain.
As a result I'm one of the reviewers of sections of the latest report on climate change and its impact by the United Nations-sponsored Intergovernmental Panel on Climate Change (IPCC), the first part of which was released last week in the form of a general summary for policymakers.
I had some serious concerns about the sections of the much bigger report that I reviewed — which hasn't yet been released---and I have some of the same reservations about the document that was published last week.
The report’s language appears to be sometimes coupled with a selective reading or oversimplification of the facts, so that the authors have "high confidence" in something that is not the whole story.
My biggest concern about the climate report is that it presents a number of speculative, and sometimes incomplete, conclusions embedded in language that gives them more scientific heft than they deserve. The report, in other words, is "scientific-sounding," rather than clearly settled and based on indisputable fact. Established facts about the global environment exist less often in science than laymen usually think.
The authors interpose a layer of opinion disguised as fact through their use of a language of "high confidence" and lesser degrees of that, and levels of “certainty” ranging from “unlikely” to “certain.”
What do these “confidence” levels and “unlikely to virtually certain” levels mean?
The report states that “confidence” is based on “mechanistic understanding, theory, models, expert judgment,” and that certainty levels can also be based just on “expert judgment.” The fact is, science doesn't work like that. I can’t think of any major scientific advances that were put in these terms as a key way to evaluate them.
Equally important, the just published summary does not show where its data came from— it refers to a “box” and “chapter 1" of the full report, which aren’t available now. This makes it all the tougher to evaluate the language of “high confidence" etc., which, in the end, papers over whether the material is truly based on science or not.
For example, the report states that “on a global scale, the ocean warming is largest near the surface.” Either this is a truism (the sun must be heating the ocean surface first) or it is meant to take into account the complex circulations that occur in the ocean, like the Gulf Stream’s involvement in a vertical rise of waters from deep ocean layers in one region and sinking of the cooled surface waters as the stream reaches its northern limit.
The report would be more meaningful if it made a straightforward statement about the measuring methods and data on which this conclusion is based. The fact that it doesn't may make it easier for non-scientists to repeat the generalization, but it obscures what this debate needs most: transparency about methods, and clarity about facts.
Indeed, the report’s language appears to be sometimes coupled with a selective reading or oversimplification of the facts, so that the authors have "high confidence" in something that is not the whole story.
For example, they state that over the past two decades, Arctic sea ice cover has “continued to decrease in extent (high confidence).” But in 2010 I co-authored a paper with two sea ice experts and a historian/ethnologist in which we used nineteenth-century records from whaling ships hunting bowhead whales in the north Pacific and beyond, along the Arctic sea ice edge.
We compared 23,000 days of observations in those records with late twentieth-century observations, and concluded that the extent of the sea ice at the end of winter was pretty much the same in the nineteenth and late twentieth century, but that the end-of-summer Arctic sea ice retreat is greater today than it was then. These kinds of more subtle findings could make a big difference in one’s level of “confidence."
The report’s heavy dependence on existing climate models creates further problems. Models, like all scientific theory, have to be tested against real-world observations.
Physicists still continually try to test Einstein’s theory of relativity as new opportunities arise, even though it was first published in 1916. The theory has cleared test after test, but researchers keep trying to find its limits and thereby also the limits of its applications. The most recent test of relativity took place in 2013.
Scientists call such tests “validation.” Experts in model validation say that the climate models frequently cited in the IPCC report are little if any validated. This means that as theory they are fundamentally scientifically unproven.
Many versions of climate models are in use, but they are all based on the same theory. We do not have competing theories, just competing variations of one theory. As a result, the current report can confuse a reader even more as to what is fact and what is speculation based on poorly validated models.
This is another way that opinion is hidden under what is called science; it is theory untested and therefore not by itself science. A more powerful science would involve strongly different ways of approaching theory, to see which accounted for more observations. And following the standard scientific rule called “Occam’s Razor,” we would go with the simplest model that accounted for the same data as a more complex one.
The bigger problem is that when it comes to climate, models have become our new reality, rather than something we test and discard when reality brings their weakness to our attention. Because global environmental data are generally so sparse and have not been gathered comprehensively for long, the tendency is to believe the models as our truth.
This will be to our disadvantage. Having worked for decades on climate change and its possible effects on life, I come away from the just published "summary for policymakers" believing that it does not allow a scientist, let alone a policymaker, to decide that we are, or are not, creating a global warming.
As a result, I foresee two dangers. One is that it will simply intensify the political, ideological and, yes, moral debate that has erupted over who does and does not believe we are causing global warming, and thus move us even farther from the important scientific effort that the issue deserves.
The second danger may be even worse: it reinforces the belief that there is some kind of climate normality, usually characterized as existing earlier in the twentieth century or before the Industrial Revolution (but after the Little Ice Age, which lasted from approximately the mid-1400s to 1700) that is desirable, even necessary, for our species and for the ecology of the planet.
In fact, there has never been such a thing, which is one of the reasons that biological evolution and adaptation exist in the first place.
As I discuss in my book "The Moon in the Nautilus Shell," only by using nature's dynamics as a guide, rather than best-guess, sweeping climate models, will we understand the naturalness of change, and thereby conserve the biodiversity of life on Earth---that is, for millions of species, including our own.
Daniel B. Botkin is Adjunct Professor of Biology, University of Miami, Coral Gables, Fla., and Professor (emeritus), University of California, Santa Barbara. His most recent books are "The Moon in the Nautilus Shell," (Oxford University Press 2012) and "Powering the Future: A Scientist’s Guide to Energy Independence" (FT Press Science). He was a designated expert reviewer of the White House Climate Change Assessment. For more, visit his website: danbotkin.com. Follow him on Twitter @danielbotkin.